dis(1)

Synopsis

   dis [-osV] [-A addr] [-c cnt] [-f fmt] [-F symb]
   	[-n name] [-O output] [-w sz] file . . .

   dis [-L] [-dDt sect] [-l lib]

Description

The following options are interpreted by the disassembler and may be specified in any order.

-A addr

Start the dump or disassembly at addr, a decimal, octal or hexadecimal number.

-c cnt

Limit the disassembly to cnt instructions or the dump to cnt data bytes. cnt is a decimal, octal or hexadecimal number.

-d section[+offset]

(Obsolete) Dump the named section as data, printing the offset as relative to the section start. If the optional +offset is present, start instead at offset bytes into the section.

-D section[+offset]

(Obsolete) Dump the named section as data, printing the virtual address as the offset. If the optional +offset is present, start instead at offset bytes into the section. offset is a decimal, octal or hexadecimal number.

-f fmt

Specify format of instruction output, where fmt is one or more of the following characters:

-

specifies the default format (abl)

a

include the instruction address

b

include the instruction raw bytes

c

specifies the compact format (just instructions) - cannot be combined with other fmt characters

f

specifies full format (ablt)

l

include source line numbers if available

t

include source text and line numbers if available

-F symbol[+offset]

Disassemble either the function symbol or dump the object symbol. If the optional +offset is present, start instead at offset bytes into the symbol. offset is a decimal, octal or hexadecimal number. By default, if symbol is found in a writable section, it is assumed not to be a function.

-l name

(Obsolete) Disassemble or dump a file named libname.a found using the default ld(1) library search path.

-L

(Obsolete) No effect.

-n section[+offset]

Disassemble or dump the named section, using the virtual address as the offset. If the optional +offset is present, start instead at offset bytes into the section. offset is a decimal, octal or hexadecimal number. By default, if section is writable its contents will be dumped as data.

-o

Print offsets in octal instead of hexadecimal.

-O output

Specify the decode model, where output is one of the following:

data

forces a data dump

text

forces a disassembly, guessing the correct instruction decoding

x86

ia32

ia32-32

forces a disassembly taking the content as Intel IA-32 instructions with 32-bit data and address sizes

286

ia32-16

forces a disassembly taking the content as Intel IA-32 instructions with 16-bit data and address sizes

-s

Append symbolic disassembly information when possible and appropriate.

-t section[+offset]

(Obsolete) Disassemble the named section as data, using the virtual address as the offset. If the optional +offset is present, start instead at offset bytes into the section. offset is a decimal, octal or hexadecimal number.

-V

Print, on standard error, the disassembler's version information.

-w sz

Specify the dump word size to be sz, a decimal, octal or hexadecimal number which must be equal to 1, 2, 4 or 8.

If given no options, dis will disassemble all nonwritable sections for each file. The -d, -D, -n or -t options restrict dis to operate only on the named section. Similarly, the -F option restricts dis to operate only on the named symbol, only one of which will be present in any object file.

On output, a number enclosed in square brackets at the beginning of a line, such as ``[5]'', indicates that the break-pointable line number starts with the following instruction. These line numbers will be printed only if the file was compiled with additional debugging information (for example, the -g option of cc). An expression such as ``<40>'' in the operand field or in the symbolic disassembly, following a relative displacement for control transfer instructions, is the computed address within the section to which control will be transferred. A function name will appear in the first column, followed by ``()'' if the object file contains a symbol table. When debugging information is present, the function name will be prefixed by file@, indicating the originating source file.

Errors

Files

Libdir

usually /usr/ccs/lib followed by /usr/lib.

References

Notices

At this time, symbolic disassembly does not take advantage of additional information available if the file is compiled with the -g option.

This command has been updated to handle Intel Pentium III Streaming SIMD instructions; see ``Pentium III extended floating point support'' in New features for more information.